Plantand Soil 176: 139-147, 1995. © 1995 KluwerAcademicPublishers. PrintedintheNetherlands.
Spatial variability of throughfall chemistry and selected soil properties as influenced by stem distance in a mature Norway spruce (Picea abies, Karst.) stand J. Seiler and E. Matzner Department of Soil Ecology, BITOK, University of Bayreuth, D-95440 Bayreuth, Germany* Received6 February 1995.Acceptedin revisedform 1 May 1995
Key words: Picea abies, soil chemistry, spatial variability, throughfall chemistry
Abstract Our aims were to investigate the spatial variability of throughfall chemistry and soil parameters as influenced by stem distance and to evaluate the implication of the observed systematic and random patterns for the sampling strategy. One hundred throughfall samplers with a sampling area of 106 cm 2 each were established in a systematic grid around 5 trees in a mature Norway spruce; site of the Fichtelgebirge (Germany). One hundred soil cores were taken with an auger of 50 cm 2 next to the throughfall samplers. Soil samples were stratified according to genetic soil horizons and analysed for pH, exchangeable NH +, SO~- and total-S. Throughfall samples were collected over a period of 6 months. For each sampler an aliquod sample was mixed over the observation period and analysed for major ions. The spatial variability of the element concentrations in throughfall, expressed by the coefficient of variance, was 21-164%, depending on the element considered. For precipitation volume, the coefficient of variance was only 3%. The distance to the stem influenced most element concentrations in throughfall with increasing concentrations approaching the stem. Steepest gradients were observed in case of SO 2- and H + . The spatial variability of the investigated soil parameters was also very high with the exception of pH. The SO42-content of the forest floor reflected the gradients observed in throughfall, while for the other investigated soil parameters and soil horizons no significant relations to stem distance were found. To determine site representative throughfall concentrations and soil properties with the sample volumes and time intervals we used, the number of samples required to get a statistical error of less than 10 % (with 95 % probability) can be very high. In case of throughfall, more than 100, and in case of the soil parameters, more than 300 replicates would be required.
Introduction The determination of site representative throughfall fluxes and soil parameters site is complicated by random and systematic variabilities on all scales considered. Knowing systematic patterns of spatial variability allows the adaptation of sampling strategies to minimize the effort of sampling and analysis and to estimate the potential statistical errors of the average values. The latter is especially critical when, e.g., comparing * FAX no: +49 921 555799
periodic soil inventories or establishing budgets using throughfall fluxes. Furthermore, the interpretation of areal average values for a specific parameter might be misleading without knowing its spatial pattern in the field. In forest ecosystems, the tree is a major structural component that may create spatial patterns of soil chemical parameters by root uptake, litter fall, throughfall and stemflow. Stemflow influence on soil chemistry is especially pronounced in beech sites (Glatzel and Kazda, 1985; Glavac and Koenies 1986; Jochheim and Sch~ifer, 1988; Falkengren-Grerup and Bj6rk,
140 1991). Also for Norway spruce, stem distance related gradients of throughfall chemistry and thus of the element input to the soil have been found. Beier et al. (1993), Freiesleben v. et al. (1986) and Pedersen (1992) report increasing concentrations in throughfall approaching the stem in 40 yrs old Norway spruce and Sitka spruce stands. The spatial distribution of needle biomass is a major factor influencing throughfall chemistry through its effect on dry deposition (Lindberg, 1992), leaching of elements from foliage (Tuckey, 1970) and by the uptake of elements into the foliage (Eilers et al., 1992). Gradients in soil solution chemistry with stem distance were reported in case of Norway spruce by Gundersen et al. (1995), Manderscheid and Matzner (1995) and Koch and Matzner (1993). Other authors claimed, that the soil solid phase close to trees has a stronger degree of acidification than the soil between stems (Norden, 1994; Pallent and Riha, 1991; Skeffington, 1983). Soil acidity patterns in these cases were related to litter accumulation. The potential relation of the observed spatial patterns in soil chemistry to throughfall chemistry was partly hypothesized, but not investigated. The aim of our study was thus to investigate stem distance related patterns of throughfall chemistry in a mature Norway spruce site and their relation to patterns of selected soil parameters. We were furthermore interested in the implication of the spatial variability for sampling strategies and for the accuracy of areal averages.
Methods
Site The investigated site is the 140 years old Norway spruce (Picea abies, Karst.) stand "Coulissenhieb"in the Fichtelgebirge (NE Bavaria, Germany). The site is about 2.5 ha in size and located in the "Waldstein"watershed at an altitude of 800 m. The stem density is 320 ha-l with trees of an average height of about 30 m. The slope of the site is less than 5 % to southwest. Ground vegetation is represented by patches of
Deschampsia flexuosa, Vaccinium myrtillus, Calamagrostis villosa or areas lacking ground vegetation. The average annual precipitation is about 1100 mm and the average annual temperature is 5.2 °C. Like most of Central European forests, the site is highly impacted by the deposition of air pollutants. Annual rates of total deposition of major elements were
Table 1. Estimated total deposition and measured average throughfall fluxes of ions in 1993 (1 Jan. - 30 Dec.) at the Coulissenhieb site (kmolc.ha-l.yr-l) Element
Total deposition
H20[mm] Na + CICa 2+
Throughfall
763 0.26 0.34 0.22
0.26 0.34 0.61
K+
0.10
0.59
Mg 2+
0.02
0.12
H+ AI 3+ Fe 3+
0.61 0.04 0.00
1.16 0.12 0.03
Mn 2+
0.00
0.03
NH4+-N
0.66
0.60
NO 3 SO2 -
1.03 1.96
0.98 1.96
HPO42-
0.06
0.02
calculated according to Matzner (1989) from measured average areal throughfall fluxes in 1993 (Manderscheid and Matzner, 1995) and are given in Table 1. The throughfall chemistry is dominated by H +, and NH4+ as major cations and by SO42-, and NO~- as major anions. The rates of total NH + and NO~ deposition given in Table 1 are underestimations because gaseous deposition of NH3 and NOx is neglected during the calculation. The soil has developed from weathered granitic bedrock. Different soil types can be found on the plot classified as random mosaics of Cambisols and Cambic Podzols (FAO). The soil texture is loamy sand to loam.Standard chemical properties of the mineral soil horizons were determined by analysing mixed soil samples from a single, large soil pit (Table 2). The soil is very acid with pH(KC1) ranging from 2.7 to 4.1 and base saturations of CEC of less than 10%. Five trees with well developed crowns reaching the upper canopy of the stand (about 30 m height) were selected for further investigations. In direction of the 4 cardinal points, throughfall collectors of 106 cm 2 were placed at distances of 60, 120, 180, 240 and 300 cm to the stem (Fig. 1). In order to exclude litter and other contamination, the throughfall samplers were equipped with filters. In total, 100 sampler were established. At the position of the throughfall samplers, soil samples up to a depth of 40 cm were taken in
141 Table 2. Average soil chemical properties Horizon
Depth
pH
pH
(cm)
(H20)
(KC1)
g(100 g - 1)
Ahe Bhs Bsv BvCv
0-9 9-24 24-38 38-54
3.5 3.8 4.2 4.3
2.7 3.5 4.1 4.2
4.8 5.7 4.0 1.0
0.2 0.2 0.2 0.1
28.1 3. l 0.6 0.2
1.5 1.4 1.3 1.4
1.3 0.8 0.8 1.0
4.2 1.7 0.7 0.4
0.9 0.5 0.2 0.1
Cv
>54
4.4
4.1
<0.2
<0.0
0.1
1. I
0.9
0.3
0.1
treen o . ! ~
C
N
H+
Na +
K+
Ca 2+
Mg 2+
Mn 2+
Fe 2+
AI 3+
0.0 0.1 0.1 0.0
6.7 11.5 0.3 0.0
88.9 152.7 69.7 38.7
0.0
0.0
25.7
(NH4C 1 exchangeable cations (/zmolc g - 1))
[treeno.2~~-------qltreeno. 3~t ~
• z/"
/J
tree no. 4 ~
tree no. 5 "¢~ ~
highcrowndensity
~
lowcrowndensity positionofsampling points
i
; i
i
; i,~
Fig. 1. Position of soil sampling points and throughfall collectors.
April 1993 with an auger of 50 cm 2 area. The soil cores were separated according to soil horizons before further processing. Throughfall samples were collected fortnightly and mixed over the sampling period (May - October 1993) on an aliquod basis for each sampler. Samples were stored at 2 °C and filtered < 0.45 #m before chemical analysis, pH was measured by a glass electrode, major cations (Na +, K +, Ca 2+, Mg 2+, Mn 2+, Fe 3+, A13+) by ICP-AES. Ammonium was determined colorimetrically in a flow injection system and major anions (SO42-, NO 3, C1-) by ion chromatography (IC). Total dissolved nitrogen was measured by a total
N analyser based on chemoluminescence of NOx. Dissolved organic N (DON) was calculated as difference between inorganic and total N in solution. Undried, field moist soil samples were homogenized and analysed for exchangeable NH4+ by 1 h extraction with 1 M KC! in a soil/solution ratio of 1:10. Ammonium in the extracts was measured by flow injection analysis. Sulfate was extracted with 0.02 M NaH2PO4 in a soil/solution ratio of 1:5 over 18 h with subsequent IC determination. We determined total-S by a HNO3 digestion and ICP-AES analysis, pH was measured by a glass electrode in H20 with a soil/solution ratio of 1:2.5.
142 Table 3. Minimum number of samples required for the determination of element concentrations in throughfall with different fixed accuracies Minimum No. of samples with required accuracy Parameter
+/- 10%
+/-20%
H+
47
12
Na +
72
18
K+ Mg 2+ Ca 2+ Mn 2+
35 45 41 44
9 11 10 11
Fe2+
(210) a
(53) a
AI3+
(10133) a
(258) a
Nto~ DON
17 23
4 6
NH +
23
6
NO 3 SO42-
31 54
8 14
143
36
1
1
C1Precipitation volume
aMost values for this parameters were dose to or under
the detectionlimit!
We used the SPSS-PC software for statistical analysis of the data.
Results
Throughfall chemistry As indicated by the coefficient of variance (CV, Fig. 2), the spatial variability af the element concentrations in throughfaU can be substantial. Lowest spatial variability was found for the H20 fluxes. Element concentrations in throughfall had CVs of 21 to 164% depending on the element considered. The very high CVs for Fe and AI are caused by element concentrations close to detection limit. The influence of stem distance on throughfall chemistry differed according to the element. With the exception of NH + and Cl- all element concentrations were significantly (p < 0,01) correlated with distance from the stem. Relatively small gradients of the concentrations were observed for DON and total N, while the concentrations of H + and SOl- in throughfall showed a pronounced gradient with increasing concentrations approaching the stem. For Na +, K +,
Mg 2+, Ca 2+ and Mn 2+ an intermediate behaviour was found. No gradient was observed in case of the water fluxes. Relatively high water fluxes (average of 404 mm throughfall) and high rainfall intensities during the summer 1993 might explain this observation. Given the data set from our 100 samplers, we were able to calculate the minimum number of throughfall samplers required to determine average areal throughfall concentrations with an accuracy of +/- 10% and +/- 20% (a = 0,05). These values for minimum sample sizes are only valid for our site, but may serve as a rough estimate for other sites as well. Calculation was done according to the Equation 1 (Sachs, 1984): nx =
(zc~ / (x -
/1)) 2 ~r2
(I)
with nx = minimum sample size; z~ = z-value (here: 1,96 for a = 0 , 0 5 ) ; x = arithmetric mean of sample;/, = arithmetric mean of population; x - # = difference of arithmetric mean of the sample and parametric mean = required accuracy; cr = parametric standard deviation, here conservatively estimated as standard deviation of the sample. The calculated number of throughfall samplers required to fulfill the chosen precision varies with the element. Considering all ions, with exception of Na + and e l - , about 50 samplers will be required for an accuracy of +/-10% under our site conditions. A minimum number of replicates for AI and Fe was difficult to define because their concentrations were close to the detection limit. In case of Na + and Cl- the variation is higher as compared to other major ions and the required number of samples exceeds 50 by far. For determining the precipitation volume on the +/-10% level only 1 sample would be sufficient.
Soil parameters To demonstrate the general spatial variability of the investigated soil parameters, arithmetic average concentrations, minima and maxima and standard deviation are given in Figure 3. Because morphologically different A and B horizons were sampled, the number of A and B horizons exceeded 100, despite only 100 cores. All parameters had pronounced depth gradients. In case of exchangeable NH +, highest amounts were found in the forest floor, decreasing almost to the detection limit in Bhorizons. Extractable SOl- is relatively low in the forest floor and A-horizon but high amounts are obtained from B-horizons. The highest total sulfur contents were
143
ICC ~ ".0,62 AM 155 CV¢~] 35 s l~ I ~z
q
= AM CV I~/ SOs)
3OO
30 ~
20'
1O0
10" o 60
120
180
240
300
60
360
distance from stem basis [cmJ
2+
CC
I
=
180
lsFq
=
240
300
360
4r
. . . . o,~1
60
8O
120
180
240
300
360
distance lram stem basis {cm]
[CC = ~,35 AM 43 CV f~] 33 40 S I~J
34
~ ~
!'i'
120
distance from stem basis [cm]
= ~-0,39
| CV¢~1
Icc
!:ii
40 1
200
,---I
"-0.30 24 42 5f
r
6o 40
u
.20 I
I
(
I
I
60 120 180 240 300 distance from stem basis [cm]
ICC I
360
0
,.
= "-0,54
CV~
=
;'4
Is~l
=
21e
60 120 180 240 300 360 distance #ore stem basis [cm/
60
120
180
:-%
300
360
I CC = "°.0.33 AM 236 CV ¢l~/ 21 ISf~l 22
400
3
240
distance from stem basis [cm/
lOO 0 0
60
120
180
240
300
360
60
d/stance from stein basis [cm]
ICC = AM CV tlg IS~
6o 2O 0
t
0
120
180
240
300
360
0
distance from stem basis/cm}
r4•
".0,28 ,115 24 23
i cAM c
=
cvt~) s f~ j
60
120
180
240
300
360
distance #ore stem basis [cmJ
_'~
.0,22 79 24
CC AM CV~) IS~l
]
= = = =
*-0,2~ 11t 2¢ 2z
120
!,o 6O 30 0
60
120
180 240
300 360
60
distance from stem basis {cm]
S~
250 !
I
120
180
240
300
360
o
distance from stem basis [cmJ
JCC AM CV(~J [ S fla/
120
= = = =
180 240
300
360
300
360
s~l
i
20O
O'
120
180 240
6O0
30'
60
120
Ic°''c,,,.,,=
-0,10 46 61 --
,oo
distance from stem Deals {cm]
60
distance #ore 8tem basis [cm]
60
120
180
240
300
distance #ore stem basis [cm]
360
o 0
60 120 180 240 300 distance from stem basis [cml
360
Fig. 2. Concentrations o f ions in throughfall in relation to stem distance (Arithmetic means as filled squares, standard deviations as bars; C C --- correlation coefficient o f concentrations a n d stem distance: * f o r p < 0. 01, ** for p < 0.001; A M = arithmetic m e a n o f all samples, C V = coefficient of variance for all samples, S= slope of the gradient).
144
I
legend
I
slanderddeviation
I
m~lum
J ForestFIoor
I
6,00
I
pH.~o o (99~[
4,00
J
I
Ah (24) l AehlAhe/Ae (76) ~
l
2,00
I
(13) ]
6,)76) /
2,8
0 "
I
I
,
,
,
,
,
3,2
3.4
3.6
3,8
=
~
j
~
ii I NH'÷
j
60
120 180 240 300 360 distance from stem basis [cm]
I
I
3
[ 1
O,O0 /
I
BsvlBvs/Bs (39) ~ Bsh/Bhs
I °
I
~sC~ = --o.~
maximum
mean value
I,
4
,
T
4,2
4.4
I I 4.6
4.8
pH(H20)
I
" content 0(99) I
j Forest ,oor =.oot
I
I
II I
i
Ah(241 "IIIIIEIT~TI~
cc = -~,4r99
SS
t
SO,?"
J
AehlAhelAe(76) Bsv/BvslB$ (39) ~--I
i %00
°,so l
)13}
BshlBhs
By (78)
0,00
1
2
3
4
5
6
7
8
9
10
11
12
O
13
soil content[mmoYkg]
I
i
f
i
,i
i
60
120
180
240
300
distance from stem basis
SO, 2" content l
Fig. 4.
360
[cmJ
Stem distance vs. soil pH, and soil N H + , total-S and SO42-
(Arithmetic meanand standarddeviation; CC = correlationcoefficient, concentrationvs. stem distance;SS = No. of total samples).
0 (99)
All (241 " ~ T I m - - ~ I AehlAhelAe (76} I ~ BsvlBvslBs 139) Bsh/Bhs
I
113)
I
'
I
?
,,
3
4
By (781
I
! I
• I. . . . . . 1
2
S
6
'2
I
8
9
.
10
11
.
.
12
.
13
14
15
s o # content [nvnot4cg]
S~,s content
[ I
O )99)
I
I
I
I
A..IAheIAa(76)~ I - ~ Bsv/Bvs/Bs (39) ~ BshlBhs
(13)~
6v (,6} | 0
t
i
i
i
i
i
i
i
i
10
20
3o
40
50
60
70
80
90
sot7confenl [mmoYkg]
Fig. 3. Spatial variabilityof selected soil properties.
found in the forest floor, while in A and B-horizons amounts were about comparable. The spatial variability of the exchangeable NH + content in forest floors and A horizons is extremely high as indicated by the standard deviation and the observed minima and maxima. The spatial variability of soil SO 2- was highest in B horizons while total S contents covered the largest range in forest floors. A significant correlation (p < 0,001) of stem distance to the soil parameters considered was only found in case of the NH + and SO42- content of the forest floor
(Fig. 4 and Table 4). The gradient of extractable SO ] in the forest floor is relatively small as compared to the one observed in throughfall. Since no gradient to stem distance was observed for the NH4+ concentration in throughfall, the soil gradient must have other reasons than reflecting the input of NH + by throughfall. Correlations (p < 0,01) are also indicated for NH + in Bh and Bsv horizons, for total-S in Bsv horizons and pH in Bv horizons. The minimum number of samples required to estimate the areal average soil content has been calculated according to Equation 1 (Table 5). It seems almost impossible to reach an average with 10% .precision (at 95% confidence) for exchangeable NH + and SO42in the A-horizons of our site. More than 300 replicates would be required to determine SO42- in A horizons and NH4+ in Bv and Bs horizons. Reducing the proposed level of accuracy to +/- 20% will reduce the required number of samples by 1/4. Because of the logarithmic scale, only one sample is necessary to determine pH(H20) of soil horizons on the +/-10% level.
145 Table 4. Correlation coefficients for selected soil parameters and stem distance (* =p<0.01; ** =p<0,001)
Horizon
Sample size
NH +
SO]-
StotaJ
pH(H20)
Forest Floor Ah Aeh,Aiae,Ae Bh Bsv Bv
99 24 65 24 33 75
**-0.36 -0.19 -0.08 *-0.50 *-0.45 -0.17
**-0.47 -0.28 -0.15 -0.28 -0.38 -0.12
-0.08 -0.24 -0.12 -0.11 *-0.42 -0.25
0.05 0.12 0.16 0.12 0.28 *0.29
Table 5. Minimum number of samples required for the determination of pH, and NH4+, SO] + and total-S-content of the soil with different fixed accuracies
Horizon
Forest Floor Ah Aeh,Ahe,Ae Bsv,Bvs,Bs Bsh,Bhs By
Required
Minimum No. of samples
accuracy
NH~-
SO~+
StotaI
+/-10% +/-20% +/-10% +/-20%
83 21 191 48
30 8 146 36
9 2 54 13
+/- 10% +/-20% +/-10% +/-20%
221 55 360 90
484 121 64 16
59 15 25 6
+/-10% +/-20% +/-10% +/-20%
130 32 334 84
109 27 79 20
33 8 26 7
Discussion As mentioned in the introduction, variability in natural systems is a question of the scale considered (Warrick and Nielsen, 1980). Thus the results on studies of variability will depend on the volume or size of the samples taken, with often increasing variability with decreasing sample volume and sampling area. In this study we used sampling areas and volumes for the collection of throughfall (106 cm 2) and soil samples (200-500 mL) that are common in ecosystem studies and our results are only valid for these volumes and sampling intervals. Our results showed a pronounced effect of the forest canopy (as expressed by stem distance) on the spatial pattern of throughfall chemistry. This is especially true for elements with a high rate of dry deposition and
~H(H20)
obviously little canopy interaction like SO]- and H +. In case of NH + and NO~ substantial rates of uptake into the foliage (Eilers et al., 1992) may interfere with dry deposition and flatten the gradient that would result from dry deposition. In comparison to the spatial distribution of throughfall chemistry in a 40 yrs old Norway spruce site of Denmark our results disagree partly: Beier et al. (1993) report strong gradients related to stem distance for all major ions with steepest gradients for NH+~ Mg 2+, Ca 2+ , H + and SO]-. Local deposition conditions with much higher deposition rates of most elements at the site in Denmark and the structure of the canopy (hight of the tree and needle mass) might explain these controverse findings and a steeper gradient seems to be typical for younger sites. Furthermore, the nutritional status of the trees will affect canopy processes like
146 leaching and N-uptake and will thereby modify spatial patterns of throughfall chemistry (Bruckner et al., 1993). Regarding the investigated soil properties, the spatial pattern of throughfall chemistry is only reflected in the extractable SO42- of the forest floor. In the mineral soil, the spatial variability of soil properties influencing the SO42- -pools and plant uptake seems to counteract the pattern of throughfall chemistry and no gradients were found. The gradient in throughfall pH was not related to soil pH in the upper soil profile. The observed gradients of exchangeable NH + in the forest floor must be caused by gradients of ammonification, nitrification or root uptake since no NH + gradient in throughfall occured. There was a small, but significant gradient of the thickness of the humus layer, which increased approaching the stem (data not shown). The exchangeable NH + -pool is a result of mineralisation and root uptake. One can expect a strong seasonal development of both processes resulting in seasonal variations of the exchangeable NH + -pool. We thus cannot be sure if the observed gradients found in April are stable throughout the year. Literature findings on soil chemical gradients related to stem distance in Norway spruce stands are somewhat contradictory: The pH gradients observed by Pallent and Riha (1991) were neither confirmed by Koch and Matzner (1993) nor by our own results. Pallent and Riha (1991) investigated a 40 yrs old site and gradients within less than 1 m distance to the stem. Obviously site specific processes of H + -production and consumption related to humus accumulation, mineralization and ion uptake as well as the general degree of soil acidification are decisive for the establishment of such pH gradients. The soil solution might be a better indicator for chemical gradients in the soil under Norway spruce. As was shown by Manderscheid and Matzner (1995) at the Coulissenhieb site and by Gundersen et al. (1995) and Koch and Matzner (1993) for other Norway spruce sites, the SO 2- concentrations of the soil solution reflect the spatial pattern of SO2on throughfall. Since no such gradient was found for extractable SO4z- in the soil, this implies a saturation of the SO42- adsorption places at the soil solid phase and a flat SO2- adsorption isotherm at these SO4zconcentrations (Alewell, 1995). Pedersen (1992) compared different methods of sampling and calculating average throughfall concentrations in a Sitka spruce site. He concluded that the average concentration can be best calculated by a regression of the concentrations to the stem distance
and from the location of the samplers. But the success of this method depends on the regularity of the crown and the steepness of the gradients. Only circular crowns with nearly the same radius (e.g. planted young coniferous stands) yield powerful correlations between stem distance and throughfall concentrations. The steeper the gradient of throughfall concentrations the more of the spatial variability is explained by stem distance. Our site is characterised by very heterogeneous crown structures (Fig. 1) and thus, in comparison to other studies (Beier et al., 1993; Freiesleben v. et al., 1986; Pedersen, 1992) by relatively flat gradients of throughfall chemistry. For that reason only 8% to a maximum of 38% (r 2 = 0,08 for NO 3 and 0,38 for SO2-) of the spatial variability of throughfall chemistry is explained by the distance to stem basis. In mature sites of Norway spruce a consideration of stem related gradients of throughfall chemistry in the sampling design and calculation of average throughfall chemistry will thus only be of little value.
Conclusions Our results generally indicate a large spatial variation of soil properties and throughfall chemistry. The spatial pattern of throughfall chemistry in Norway spruce sites are related to stem distance with increasing concentration approaching the stem and these gradients are specific for each element. From literature findings one can conclude that gradients of throughfall chemistry seem to be less developed in mature stands as compared to younger ones. Given our sampling volumes, the number of sampies required to yield an arithmetic, site representative, average of throughfall concentrations and soil parameters with precision of+/- 10 % can be more than 100.
Acknowledgement This work was funded by the German Federal Ministry of Research and Technology (BMFT) research grant No BEO 51-0339476 A.
References Alewell C 1995 Sulfat-Dynamik in sauren Waldb6den - Sorptionsverhalten und Prognose bei nachlassenden Depositionen. Bayreuther Forum Okologie, University of Bayreuth (In press).
147 Beier C, Hansen K and Gundersen P 1993 Spatial variability of throughfall fluxes in a spruce forest. Environ. Pollut. 81, 257267. Bruckner G, Gebauer G and Schulze E D 1993 Uptake of 15NH3 by Picea abies in closed chamber experiments. Isotopenpraxis Environ. Health Stud. 29, 71-76. Eilers G, Brumme R and Matzner E 1992 Above-ground N-uptake from wet deposition by Norway spruce (Picea albies Karst.). For. Ecol. Manage. 51,239-249. Falkengren-Grerup U and Bj/Srk L 1991 Reversibility of stemflowinduced soil acidification in Swedish beech forest. Environ. Pollut. 74, 31-37. Freiesleben v. N E, Ridder C and Rasmussen L 1986 Patterns of acid deposition to a Danish spruce forest. Water Air Soil Pollut. 30, 153-141. Glatzel G and Kazda M 1985 Wachstum und Mineralstoffern~ihrung von Buche (Fagus sylvatica) und Spitzahorn (Acerplatanoides) auf versauertem und schwermetallbelastetem Bodenmaterial aus dem Einsickerungsbereich von Stammablaufwasser in Buchenw~ldern. Z. Pflanzenern~hr. Bodenkd, 148, 429-438. Glavac V and Koenies H 1986 Kleinrtiumige Verteilung der pflanzenaufnehmbaren Mineralstoffe in den vom Stammablaufwasser beeinflu/3ten Bodenbereichen alter Buchen verschiedener Waldgesellschaften. Dtisseldorfer Geobotanische Kolloquien 3, 3-13. Gundersen P, Andersen B R, Beier C and Rasmussen L 1995 Experimental manipulation of water and nutrient input to a Norway spruce plantation at Klosterhede, Denmark 1) Unintended physical and chemical changes by roof experiments. Plant and Soil 168/169, 601-611. Jochheim H and Sch~ifer H 1988 Die "Baumfu/3-Methode", dargestellt anhand einer Untersuchung der Immissionsbelastung von Nordwest-Jugoslawischen Buchenw~ildern. Z. Pflanzenerntihr. Bodenkd. 151, 81-85.
Koch A S and Matzner E 1993 Variability of soil and soil solution chemistry under Norway spruce (Picea abies Karst.) and European beech (Fagus silvatica 1.) as influenced by distance from stem basis. Plant and Soil 151,227-237. Lindberg S E 1992 Atmospheric deposition and canopy interactions of sulfur. Atmospheric deposition and forest nutrient cycling, bz Ecological Studies 91. Eds. D W Johnson and S E Lindberg. pp 74-90. Springer Verlag, New York, USA. Manderscheid B and Matzner E 1995 Spatial and temporal variability of soil solution chemistry and ion fluxes through the soil in a mature Norway Spruce (Picea abies (L.) Karst.) stand. Biogeochemistry (In press). Matzner E 1989 Acidic precipitation: Case study Soiling. h~ Acidic Precipitation. Vol. 1: Case Studies. Eds. D C Adriano and M Havas. pp 39-81. Springer Verlag, New York, USA. Norden U 1994 Influence of broad-leaved tree species on pH and organic matter content of forest topsoils in Scania, South Sweden. Scand. J. For. Res. 9, I-8. Pallent E and Riha S J 1991 Surface soils acidification under red pine and Norway spruce. Soil Sci. Soc. Am. J. 54, 1124-1130. Pedersen L B 1992 Throughfall chemistry of sitka spruce stands as influenced by tree spacing. Scand. J. For. Res. 7, 433-444. Sachs L 1984 Angewandte Statistik, 6. Auflage. Springer Verlag, Berlin, Germany. 688 p. Skeffington R A 1983 Soil properties under three species of tree in Southern England in relation to acid deposition in throughfall. In Effects of Accumulation of Air Pollution in Forest Ecosystems. Eds. B Ulrich and J Pankrath. pp 219-231. D Reidel Publishing Company, Dordrecbt, the Netherlands. Tuckey H B jr 1970 The leaching of substances from plants~ Arm. Rev. Plant Physiol. 21,305-324. Warrick A W and Nielsen D R 1980 Spatial variability of soil physical properties in the field, h~ Application of Soil Physics. Ed. D Hillel. pp 319-344. Acad. Press, New York, USA.
Section editor: R F Huettl
